|
FACTS AND FIGURES, surveys and statistics. Day after day, an ocean of data swirls about us, but in the barrage of numbers, we rely on others to interpret. What we don't understand, we simply
accept. It doesn't have to be this way, and here are a couple of evangelists to tell us why. Neither of these books will take more than a couple of evenings to dispatch, and both are refreshingly subversive. Innumeracy
by John Allen Paulos
IT'S A FACT that more Americans die of heart disease or
cancer than anything else, but more potential years of life are lost to traffic accidents. Should Americans stop smoking or stop driving? If a million seconds is eleven and a half days, how long is a billion? And just
how real is the AIDS 'epidemic'? If your reaction to such questions is to shrug and move on, take comfort in this book and know you're not alone. Paulos's premise is that most of us have little feel for the numbers
that bombard us every day. His aim is to educate us by demonstrating how little we really know and pointing out the day to day consequences of such ignorance. Urging us to look at numbers in unusual ways, he illustrates
his points with intriguing examples and anecdotes drawn from real life. Probability and statistics are a recurrent theme, and he includes a fascinating chapter on coincidence which shows just how
counter-intuitive probability theory can be. Paulos traces the roots of innumeracy to uninspired and often incompetent schooling, particularly in the formative years. What he doesn't say is that in a pocket calculator
society, atrophy of those
barely-grasped and never-exercised intellectual skills is virtually inevitable. In equating innumeracy with illiteracy, however, Paulos strains credibility: in every way we must read to
live, but getting by without numbers just isn't that big a deal. That's not to say he's wasting our time, because a dab of numerical know-how goes a long way in the crusade against corporate fibbers. The book is often
surprising and sometimes even startling, but where it stumbles is in the author's attitude to his readers. Paulos's 'innumerate' is gullible and foolish but at times he seems to forget that many readers are those very
fools seeking salvation. He is a little too scathing in his often perfunctory dismissal of 'pseudo sciences', and occasionally downright patronising. Fortunately this doesn't stop the message getting through. On the
whole it's an entertaining, eye-opening and authoritative romp down some disturbingly familiar corridors. How to Lie with Statistics
by Darrel Huff
THE BITE-SIZED 'pearls of wisdom' genre is a surprising niche in the publishing marketplace. These books seem to break all the rules: too
short to say very much, relatively high cover prices and a limited audience, yet every now and then a tiny book like Strunk and White's 'The Elements of Style', Blanchard and Johnson's 'The One Minute Manager' or
Richard Bach's 'Jonathon Livingston Seagull' hits the spot and attracts a cult following. In this tradition, Darrel Huff's subversive exposition of statistical chicanery was first published in 1954 and is still going
strong.
It's basically an awareness book - in covering the bare bones of sampling and presentation techniques, Huff prepares the ground for his real point: by choosing the right presentation, we can use a set of
figures to show just about anything we like. With genuine examples gathered from industry, he shatters our illusions of corporate honesty and ensures we'll never trust an official report again.
Readers of an
innumerate disposition need have no fear, for Huff is more concerned with human motives than the nuts and bolts. There isn't
single overt calculation in the book, and when a quick sum is required, Huff deviously
sneaks it past by talking it out in words. Shameless manipulation and underhand advertising are his targets, and having sown the seeds of rebellion he rounds off with a practical five point plan for talking back to a
statistic.
It's all simple stuff, but none the less surprising for that. Some of the examples are dated (the last update was in 1973), but the writing is easy and pleasantly ironic. If you liked the look of
Paulos's book but were afraid of all those nasty numbers, maybe this is the one to try. |
|
